One of the most urgent needs in toxicology is for methods which can measure in animals low level adverse effects of environmental agents and whose results can be more confidently extrapolated to human populations. As a new test system to contribute to this need, it is proposed that self-sustaining populations of mice be employed as test objects instead of conventional groups of animals. A pilot study in which four mouse populations have been established and maintained for several years under conditions designed to facilitate their use in toxicological investigations demonstrated the feasibility of the approach. Stable populations and parameters for their characterization having been established, it is proposed that such populations be challenged with several levels of DDT in the diet while similar populations would serve as controls. The study is proposed as a test of the utility of this new approach in toxicology. DDT was chosen as the first environmental toxicant because of the relative wealth of published toxicological information on this pesticide for many species of animals, including man. Social stresses simultaneously imposed on the mice might have a significant effect on toxicity of environmental agents; comparison of results of the proposed studies with those in the literature may give insight into this relationship.